Tobacco Documents Online

jmoking Cessatinn and \L,rcaliry Trtnds 823 APPENDIX TABLE 5. Relative risk of death (RR and 95% CI) by length of time quit as of entry date (195411957) for al1 former cigarette smokers (n = 43,559) compared with those who never smoked regularly (n = 59,351) among U.S. veteran males of all ages for selected causes of death Cause of death All former smokers All causes Allcancer BC, esu, or ca Lung cancer Other cancer AIlCVD Coronary HD Stroke Bron, emp, asthma <5 n=15,196 1.61 (1.52-1.71) 159311 1-76 (1-54-2.25) [1100/ 3.47 (1.84-6.56) [401 10.16 (7.19-14.36) [1521 1.15 (0.97-L36) [9081 157(L46-L.69)[38931 1.56 (1-42-1.70) [26641 1.36 (1.07-1.73) [440) 14.67 (8.87-24.27) [82] RR by length of time quit in years 5-9 n = 9,348 138 (1.25-1.45) 153771 1.62 (1.38-1.89) [10001 2.08 (0Q89-4.5(,) [301 6.36 (4.27-9.47) [971 1.28(L06-133)[8731 L35 (1.24-L48) [35571 1.35 (1.2t-I50) [24251 1.15 (0.87-L51 ) (4121 7.55 (4.25-13.43) (471 10-14 n = 5,735 L33(L?2-1.45)[50471 1.42 (1.16-1.73) [9191 0.45 (0.06-3.31) [24] 3.28 (L86-5.75) [66] 1.32 (L07-L64) [8291 1.31 (1.I8-L46) [3348( 1.39 (1.23-1.58) [2296] 1.19 (0.87-L.64) [396) 3.49 (1.56-7.30) (311 ?15+ n = 13,280 1.05 (J.99-1.12) [56361 1.20 (1.D4-1.39 ) [10471 1.33 (0.59-2.99) [3l1 1.77 (1,01-2.92) [72] 1.16 (1.00-I.35) (9441 L01 (0.93-1.10) (37I8[ 0.99 (0.90-L09) (2516 0.99 (0.73-1.24) (445/ 235 (1,2 1-i.58) 1371 APPENDIX TABLE 6. Relative risk of death (RR and 95% Cl) by decade among all 19i4-1957 current cigarette smokers (n = 97,518) compared with those who never smoked regularly (n = 59,351) among U.S- veteran males of all ages for selected cause of death Cause of death All cau5ls All cancer BC, csu, ur ca Lung cancer Other cancer All CVD Coronary HD Stroke Bron, emp, asthma 1954-1959 L63 (1.57-1.69) [14,2311 1.97 (l-8'1-2.14) [29431 4.92 (3.17-7.65) [171] 10.29 (7.72-13.71) [7361 1.32 (1.20-1.44) [20361 1.55 (1.49-1.62) [90951 1.55 (1.47-1.63) [62501 L27 (1.12-L46) [9321 6.24 (4.04-9.63) [2051 RR by decade of follow-up 1960-1969 1.69 (1.65-1.72) [41,2541 2.11 (2.01-2.22) [79731 6.10 (4.48-830) [4001 10.95 (9.24-12.98) (21471 1.38 (1.30-L.46) [34311 1.51 (1.47-1.55) [?4,8771 1.51 (1.46-1.56) [59181 1_29 (L21-L38) [3i71( 10.11 (8.24-12.40) [13131 1970-1979 13'_ ( L49-1.95) [45,086] L90 ( L82-2_00) [8206] 4.C0 (3.09-5i.19) [3881 1 ~`.85 (9.17-12.84) L19461 1-30 (1 24-1.37) [58721 1.33 (1.30-137) [26,0591 L31 (1.27-136) (15,2381 L.?0 ( L L3-1.27) [5266 11 72 (930-14-78) [9981 1954-1979 1.60 (1.53-L62) [100,571) 2.00 (1.94-2.06) [19,12 i1 4.88 (4.07-i.84 ) (9591 10.86 (9.73-12.13) ['48291 1.34 (l.'_9-1.38) [13,3391 1.43 (1.41-1.46) [60,0311 1.43 (1 .4tL1.46) [37,4061 1.24 (1.19-1.29) [99691 10.28 (8.9C-11.88) 125161 APPENDIX TABLE i. Relative risk of death (RR and 95% CI) by decade among all 1954/57 current cigarette smokers by num- ber ber of cigarette per day compared with those who never smoked regularly among U.S. veteran males of all ages for all causes _ and lung cancer. Adjusted total is a weighted average based on the smoking distribution as of 195-V/57 Cause of death AIl causes Adjusced total <l cYg/dav (n = 4,115) 1-9 cigs/day (n = 13.147) 10-20 cigslday (n = 45,492) 21-39 cigs/day (n=28,932) ?40 cigs/day (n = 4.932) Lungcancer Adjusted total <I cig/day 1-9 cigs/day 10-20 cigs/day 21-39 cigs/day y40 cigs/day I954-1959 1 63 (1.57-1.69) L14 (1:02-1.28) 147911 1 23 (L16-L.31) [i6741 1.61 (1.54-L68) [9C461 1.92 ( 1 .34-2.0. ) 175091 2.19 ('-.01-2.33) [50951 10.3 (7.7-13.7) 3.53 (1.84-6.77) [611 4.39 (2.96-6.53) [981 8.28 (6.11-11.21) [3111 16.85 (12_44-2282) [338J 25.10 ( 17.47-36.07) (128] RR by decade of follow-up 1960-1969 1]0(I_66-L73) 1.16 (1-09-1 2 4) [14_'631 L.30 (L25-L35) [I "0.9141 1.66 (L.62-1.7i) 176,3341 2D 1 (1.96-2.07 ) 121,8831 2.39 ( 2.28-2.91) (15.0791 1970-1979 1.55 (1.5 2-1.56 ) I.I'_ ( L06-1-19) [19.3041 1.19 (L15-1.23) [3l.>491 1.51 (1-45-1.55) 130.5561 I 36 ( I.31-I .91) 1=5.6611 2_03 ( L.92-2.15) 118,9921 10.9 (9.2-12.9) 3.60 (2.47-5.26) (1761 3.77 (Z.95-4.83) [256] 10.03 (8.40-11-98) [1009I 16.82 (14.05-20.14) [9391 23.47 (18.83-29.25) [339( 11,1 (9.5-13.7,) 3.44 (2.34-5.05)[1801 4.12 (3.23-5.26) [2631 10.26 (8.59-12.24) [9411 17-50 (14.61-20.96) [8671 21.03 (16.56-26.71) (287j 195J-1979 L62(L6:-1.64) 1_14 ( 1_29-1-19) [37,s631 1 24 0 -'_1-L1-7) 144,1371 158 (1.56-L.6U [65,9361 1.93 (1 ~C-1.97) (55,0531 2.21 (2.14-'_'9) (39,1661 1 1.0 (9.E-1=.2 ) 3.54 ('-76-4.55) [4171 4.02 ( 3.43-4.71) [6171 R92 (8-94-1 L.14) (22611 17.19 (15.'_3-19.33) [21441 22.75 (19.63-26.37) 1754[

814 smokers the longer the former smokers have noc smoked [4[. Among former smokers, whose smoking atatus was de- termined at the time they entered an epidemiological study, the decline in risk of death compared with never-smokers begins during the first 5 years after quitting and continues for at least 10-15 years. After 15 years, the risk of all-cause mortality returns nearly to that of never-smokers and the risk of lung cancer mortality drops to about twice that of never-smokers. The observational studies have the limitation that the reported benefits of cessation are based on mortality pao- terns among persons who were already former smokers at the time they enrolled and were classified by the number of years since they last smoked at time of enrollment. The self- selected former smokers who enrolled in these studies are those who were alive at rhe time the study began. A former smoker who stopped 15 years before the study began is by definition one who remained alive those 15 years. Former smokers who died before the study began would obviously not be included and might be different than the self-selected former smokers who were included. The mosc rigorous tvay to evaluate cessation is to ran- domly assign smokers to either a cessation intervention or no intervencion in a randomized controlled trial (RGT). There has been just one completed RCT designed to evalu- ate smoking cessation alone: the Whitehall Civil Servants Study of 1445 middle-aged whire men in London [5J. There has been one RCT in which smoking cessation was the ma- jor risk factor change during 6 years of intervention: the Multiple Risk Factor Incercention Trial (MRFIT) of 12,866 middle-aged white men in the United States (61. Initially, both studies showed substantially more smoking cessation in the intervention group relative to the control group (about 45% versus 20% in MRFIT averaged over 6 years and about 60% versus 254o in Whitehall averaged over 9 years), but the cessation differences diminished substan- tially over time. For the intervention groups relative to their respective control groups, the total mortality during 16-20 years of follott-up was 6% less in the MRFIT and 7% less in Whitehall; lung cancer mortality was 15 :b more in MRFIT and 10% less in Whitehall. None of these differ- ences are statistically significant. The ongoing Lung Health Study, designed to evaluate smoking cessation among 3, 702 men and 2.185 women, shotts no differences in lung cancer or total mortalicy durin_ the first 5 years 171. So the RCTs have not definitivek confirmed the value ofsmokrng cessa cion, but they are limited by the fact chat they were not able to sustain a large difference in smoking cessation between intervention and conaol groups- We provide additional insight into this issue b% evaluat- ing natural experiments of smoking cessation among two U.S. cohorts, whereby the smokers as a whole largely quit, and consequently their smoking-related death races should converge toward the corresponding death rates among those who have never smoked. A natural experiment ap- J. E. Ensmtnt prosintates a concrolled experimenc, and inferences abouc etiological factots derived from such situations are consid- erably stronger than inferences derived solely front an ob- servational study, but they are not as strong as those drawn from an RCT [8[. The study of 6ritish physicians is a prime example of a natural experiment (9J. The British study in- volved 34,440 physicians who substantially reduced their cigarette smoking from 1951 to 1971 and whose lung can- cer death race declined relative to the general population rate and converged toward that of nonsmokers during these 20 years. The major mortality benefits of cessation were among smoking-related causes in physicians less than age 6) at death, with no overall benefits for ages 65 and older (9[. 1954-79 U.S. VETERANS STUDY METHODS This is a studv of U.S. veterans who held Go.'ernment Life Insurance Policies, primarily white men tcho served in World War 1(WW I) [10-131. A questionnaire requesting informatinn about the use of tobacco, residence, usual oc- cupation. and industrv of employment was mailed to about 295.t'Ct' eligible subjects in January 1954. After a second mailin~ in January [957, a total of 248,046 (8-f°.6) policy holders, ased 30-84 years, responded There were _'C0,428 veterans who ever or never smoked cigarettes regularly. Of thesr, 136.302 (68%) were born during 1890-1899, essen- tially all of tt'hom were W W I veterans. They tcere followed up for survival from Januarq L. 19~-{ through September 30, 1980, using Veterans Administra- tion (VA) records, as described elsewhere (10-13]. Causes of death wue obtained from VA claims folders or front stare tital records. The underlying cause of death was coded from the death certificates for 97% of all deaths by trained nosol- ogisrs actordin, to the International Classification of Dis- eases, Se, enrh Revision (ICD7). These veteran cohort dara are on a publically available data file maintained by the Na- tional Cancer Institute [13[. Current cigaretce smokers have been defined as all te[er- ans currenrk smokim, ci,,arettes at time of enrollnxnt in 1954 or 195 i, re, udless of whether they also smoked cigars andior piprs. Veterans rcho never smoked regul;trly are chose it tuuc of enrollment, had never smoked ciaa- rettet, ciQac. or pipes or h:rd smoked them onk occa>ion- alk. Former cigarette smokers include all cigarette smokers ttho had quic smoknt, dgarectes at time of enrollmenc, re- gardles> of their cwar artd/or pipe smoking history. These definicions are consi.stent with rhose used in most pre, ious analvses [1'-I_'I but are slightly different than those used recentlq (131 - We have amalyced the U.S. Veterans cohort using Cox proporriunal h;cnrds regression to determine the relative risk (RR) of death and 95% confidence interval (CI) over time for persons who smoked cigarettes at enrollment com- F_ I

f 5moking Cessation and Mortality Trends The author thanks the National Cancer I nstittrte and the Narional Crn- ter for Health Statistics for making available to me the data files ttpon which this research is based. This research has been supported in part by a grant from The Courutl for Tobaaro Research - U.S.A:. Inc. References 1. McGinnis JM, Foege WH. Actual causes of death in the . United Srates. JAMA 1993; 270: 2207-2212. 2. Centers for Disease Control. Surveillance for selected to- bacco-use behaviors-United Sratea. MMWR 1994; 43: 1-43. 3. Landis SH, Murray T, Bolden S, Wingo PA. Cancer Statis- tics, 1999. CA Cancer] Clin 1999; 49: 8-31. 4. The Health Benefits of Smoking Cessarion-A Report of rhe Surgeon General, Rockville, MD: Public Health Service; 1990. DHHS Publication No. (CDC) 90-8416. 5. Rose GA, Colwell L. Randomized controlled aial of anti- smoking advice: Final (20 year) results. J Epidemiol Commu- nity Health 1992; 46: 75-77. 6- The MRFIT Research Group. Morralicy afrer 16 rears for par- ticipants randomized to the Multiple Risk Factor Intervention Trial (MRFIT). Circulation [996; 94:946-951. 7. Anthonisen NR, Connett JE, Kiley JP, Ahose MD, Bailey WC. Buist AS, et aI. Effects of smoking intervention and the use of an inhaled anticholinergic bronchodilator on the Tate of decline of FEV 1(The Lung Health Study). JAMA 1994; 272:1497-1505. 8. Lilienfeld AM, Lilienfeld DE. Foundations of Epidemiology. New York:Osford University Press, 19SQ pp. 10-1 l. 9. Doll R, Peto R. Mortality in relation to smoking: '_0 years' obser- vations on male British doctors. Br Med ] 1976; 2- 1525-1536. 10. Kahn HA. The Dom study of smoking and mortality among U.S. veterans: Report on eight and one-half qears of observa- tion. Nad Cancer Inst Monogr 1966; 19: 1-125. l1. Rogot E, Murray JL. Smoking and causes of death among U.S. Veterans: 16 years of observacion. Public Health Rep 1980; 95: Z13-2??. 12.. Rogot E, Hrubec Z. Trends in mortnlipfrom curonan• heart disease and stroke amone U.S. Veterans: 1954-79. J CIEn Epi- demiol 1989; 42: 24i-256. 3. McLaughlin JK, Hrubec Z. Biot WJ, Fraumeni JF Jr. Smoking L"and cancer mortality among U.S. veterans: a'_6-,ear follorc- up.IntJ Cancer 1995; 60: 190-193. - 825 l4. Harrell F. The PHREG procedures. In: SUGI Supplemental Library Uscr's Guide, 1989 Edition. Carn. NC: SAS lnsti- tute; 1989. 15. Haens;el W, Shimkin MB, Miller HP. Tobacco smoking pat- terns in the United States. Washington. DC: U.S. Govem- ment Printing Office; 1956, Public Health Monograph No. 45. l6. Erucrom JE, Godlev FH. Cancer mortality among a represen- tatice sample of nonsmokers in the United States during 1966-68. J Natl Cancer Inst 1980; 65: l 175-1183. 17. U.S. Bureau of the Census. Current Population Survey, Sep- cember 1985. Washington, DC. l8. Giocino GA, Schooley MW, Zhu BP, Chrisman JH, Tomar SL, Peddicord JP, et al. Surveillance for selected tobacco-use bchariors-United State>, 1900-1994. MMYVR 1994: 43/Nu. SS-3: 1-f3. 19. Shopland DR. Tobacco use and its contribution to earlv can- cer mortality with a special emphasis on cigarette smoking. Environ Health Perspect 1995; 103 (Suppl 8): 131-142. 20. Bums DIM, Lee L, Shen LZ. Gilpin E, Tullec HD, Vaughn J, et a!. Cigarette smoking in the United Stares. In: Smoking and Tobacco Control .Lfonograph 8. Changes in Cigarette-Relared Disease Risks and Their Implication for Prevention and Control. National Cancer Institure; 1997: 13-11'. NIH Publ Na- 97-4213. Bethesda. MD. 21. Enstrom JE Kanim LE, Klein MA. Vitamin C intake and morralitp among a sample of the United States population. Epidemiology 1992; 3: 194-20?. 22. Natiunal Center for Health Statistics. Plan and Operation ot the NHANES I Epidemiologic Followup Study, 1982-84. Hvattsille, MD= Vital and Health Sratistics; 1987. Series 1. Number 22. DHHS Publication No. (PHS) S7-1324. '3. Nationat Center for Health Smtistics. Public Use Data Tape Documentation. NHANES I Epidemiologic Followup Study. 1992: Viral and Tracing Status. Hyatrsville, V1D: NCHS; 1996. 24. Enstrom JE, Heath CW Jr. Smoking cessation and mortality trends among 118,000 Californians, 1960-97. Epidemiology 1999;10:999. 25. National Center for Health Statistics. 1986 National Mortal- in' Folluoback Sunev. Viral and Health Statistics Series 1, No. 29. DHHS Pub. No. (PHS) 93-1305. Public Health Ser- cice. Hqact>ville. MD. May 1993 and unpublished tables. 26. Thun YIJ, Day-Lallp CA, Calle EE, Flanders lL'D, Heath CW, Jr. Excess morrnlity among cigarette smokers, Changes in a 20-ce:v intercal. Am J Public Health 1993; 85: 1223-1230.